So, the first thing you have to realize is that the 10' C-band antennas are pretty solid, heavy and with a fair degree of wind resistance. As the reflecting surface is typically expanded metal with 2 mm perforations, this thing is going to present quite a wind load. Even after removing all the C-band electronics and unnecessary aiming motor, the parabolic and its mounting structure is probably 150 lbs...so this is not going up on a mast.
What it does get mounted on is 3" galvanized steel, and for obvious reasons that pipe needs to be at least 5' out of the ground. In my case, I went for 7 feet out of the ground...but that means 5' in the ground, surrounded by at least 300 lbs of concrete. Unless you have a really big augur, it's hard to do much more than that. Luckily, the ground drops off significantly from where the pole is, so it's effectively maybe 12' AGL.
Once the dish is mounted, configuring the antenna element is pretty straightforward. I live in an area with almost no signals in any direction, so I don't need a reflector behind the antenna element: it's just a matter of making the element (bow-tie) the right size and in the focus of the dish.
According to theory, the bow-tie is of infinite size...and I'd like to get down to to channel 7 so the wavelength is almost 2 meters. But the focus of the dish is probably less than a meter across, so there's not going to be any advantage to an bow-tie even that big. My first pass is with bow-tie wires 16" long. (I will experiment with solid bow tie and other variations in a week or so.) Through experimentation and measurement, I find the best distance for the bow-tie element and we're off to the races.
My main signal challenges are about 100 miles away, and my baseline for comparison metrics is a pair of yagis, with the bottom one about 12' AGL actual...but due to drop off of ground level effectively 17' AGL. According to the models, I shouldn't be getting much of anything...but with the stacked yagis and a nice preamp most of the time I get about 75 virtual channels (from about 20 transmitters). The new dish with a preamp gets nearly all the channels the same way, and the first pass of spectrum analysis (using the Nuts about Nets USB-based system on a PC) shows nearly identical signal strength. The dish is more directional than the Yagis (no surprise) but not as super-directional as I was worried about.
So now: how do I make the bow-tie outperform the Yagis? They are signficantly farther off the ground, which gives them a big gain advantage (I'm guessing 6 dB)...but the "collecting surface" of the dish should give the bow-tie even more than that.
The answers come next time I have time to mess with it.